The broad long-term objectives of this project are (1) to improve the long-term clinical outcome of patients with proximal deep-vein thrombosis (popliteal, femoral or iliac vein thrombosis) by improving their survival and reducing morbidity from recurrent thromboembolic events, and (2) to determine if one or more specific biochemical measures, including antiphospholipid antibodies and activated factor VII, will define a subset of patients at increased risk of recurrent venous thromboembolism. The specific aim is to perform a randomized clinical trial to test the hypothesis, that in patients with a first episode of proximal-vein thrombosis, continuing warfarin treatment for 3 years will be effective for reducing mortality, and for preventing recurrent venous thromboembolism, with a low risk of major bleeding, by comparison with a control group in whom warfarin is discontinued at 3 months according to current clinical practice. This study will be performed by 13 participating hospitals in Oklahoma. 600 hundred patients will be recruited and followed for at least 3 years. The primary outcome measure will be total mortality. Secondary outcome measures will include vascular death, objectively documented recurrent venous thromboembolism, and bleeding complications (major and minor). Additional specific aims will be: (a) to determine the prevalence of antiphospholipid antibodies, antithrombomodulin antibodies, and activated protein C resistance in consecutive patients with proximal-vein thrombosis, (b) to determine if these measures, and/or the level of activated factor VII are associated with an increased risk of recurrent venous thromboembolic events, and if so, to determine if these events are reduced by continuing longterm warfarin treatment. These latter specific aims will be achieved by performing the laboratory parameters concurrently in patients entered into the clinical trial, and relating these measures to the clinical outcomes on long-term follow-up. There are more than 250,000 patients with venous thromboembolism each year in the United States. If our clinical hypothesis is correct, and continuing warfarin for 3 years reduces mortality by an absolute 10% (from 20% to 10% or less), then this would result in more than 20,000 lives saved each year in the United States. This project should provide 'explanatory' data on the etiology and mechanisms of recurrent thromboembolism, in addition to "management" data based on clinical outcomes for the practicing clinician.